Accessibility for parents of weight management information on the Internet.

Abstract

Objective: The aim of the present study was to describe the accessibility, quality and readability of web base content on the Internet for childhood obesity management from a parental viewpoint. As more than 25% of Australian children are overweight or obese, it is likely that parents access the Internet for advice on obesity management for their child.

Design: Links to websites were generated using three search engines and three phrases. The first 10 website links were selected from each of nine searches. Ten aspects of obesity management were identified for the analysis of childhood obesity and related websites. The Fry readability graph was used to assess content readability.

Data analysis: The data were analysed using SPSS Version 10, Minitab and Microsoft Excel 2000. ANOVA was used to compare clinical criteria between website type.

Results: A total of 90 links were generated with 60 relevant for analysis. In total, 30 were specific to childhood obesity and 47 websites were either obesity-specific or weight-related. Childhood obesity websites were community-based or commercial. Most websites relevant to childhood obesity management came at no cost, in comparison with general weight-related sites. Childhood obesity websites scored generally high on accuracy of content. Overall, website readability was at an undergraduate level.

Conclusions: There appears to be a paucity of accessible information on the Internet related to childhood weight management, which parents could access. Childhood obesity management websites should ideally include clinically approved management guidelines and follow e-health codes of ethical practice.

Key words: child, Internet, management, obesity, parent.

INTRODUCTION

Childhood obesity is a growing public health problem, (1) with approximately one quarter of Australian children overweight or obese. (2) A comparison of data from the NSW Schools Fitness and Physical Activity Survey, The National Nutrition Survey and the Health of Young Victorians Study found both similar weight status of Australian children (2) and that a significant increase in prevalence had occurred from 1985. (3)

Childhood and adolescent obesity increase the risk of becoming an obese adult. (4) There are significant health risks associated with childhood and adolescent obesity, including an increase in future morbidity and mortality, greater cardiovascular risk factors, type 2 diabetes, hepatic steatosis and obstructive sleep apnoea. (1) In addition, obesity affects the psychosocial wellbeing of a child. (5)

Conventional management strategies for childhood obesity include dietary modification to decrease energy intake, increased energy expenditure through increased physical activity and a decrease in sedentary activity, parental and family participation and behaviour modification. The evidence for efficacy of such treatments is limited. (6) The resources to manage established obesity in children and adolescents are also few in Australia and are often located in tertiary care facilities.

The Internet provides access to a vast body of information and overcomes geographical isolation, particularly relevant for Australia. (7) It is known that more than one quarter of Americans search for health-related information on the Internet each year, and more than 70% report that decisions about therapeutic interventions were influenced by health information that they located on the Internet. (8) A consumer's ability to assess the quality of health information on the Internet is likely to vary. (9-11) Ethical issues arising with Internet use have also emerged, with guidelines being developed. (12)

It was hypothesised that parents may seek Internet information or advice on managing obesity in their children, before seeking professional assistance or because of a lack of accessible services. The aims of the present study were: (i) to determine what childhood obesity management websites appear under the specified search conditions of the project; (ii) to review the information on the identified websites for health content, accuracy, relevance and usefulness; and (iii) to describe the content of the identified websites for parental readability and access.

METHODS

Search engines as opposed to search directories were used to minimise any variables that might influence the samples chosen for analysis. A search engine provides a list with direct links to websites, and with no options for selecting a specific area of interest (SearchEngines.com How credible is the information from the Internet?: http://www.searchengines.com/credibleInfo2.html; SearchEngines.com Search Engines: The Basics: http://www.searchengines.com/searchBasicsl.html). Search engines are an indexed database of resources obtained from the Internet by robots, spiders or crawlers that follow hyperlinks from one web document to the next to identify and index new sites. Search words are matched to the resources databases and each database is structured to filter and rank sites by relevance. Users do not have access to what procedures are used to filter and rank sites. The filtering and ranking procedure is based on the frequency of the queried word in the site, the location of terms in the document, the number of pages in the site that contain the search term and the site's design. Search engines may also take into account external influences including the site's link popularity, click popularity, sector popularity and pay-for-placement ranking.

Search engines AltaVista, Excite and Google were identified from Statmarket as the current most popular search engines in searching for web information (SearchEngineWatch.com StatMarket Search Engine Ratings. 2000. http://www.searchenginewatch.com/reports/statmarket.html). Yahoo was also rated as one of the most popular search vehicles, but as it operates as a directory, it was not used as a search vehicle for the study. In contrast, Excite involuntarily operated as an engine, although if a specific option is selected it could also operate as a directory (Monash Information Services. The Spiders Apprentice. How to use web search engines. http://www.monash.com/spida.html).

Search phrase 'weight loss and children', 'weight management and children' and 'weight treatment and children' were used, as in the authors' experience (KS) these terms are commonly used by parents in a clinical setting. The term 'obesity' was not selected as a search word as parents do not generally identify their child as being obese in a clinical weight management setting. Replacing weight by overweight or by obesity yielded similar results to the search phrases used in the present study (data not shown). 'Australia' as a search word was not selected initially so as not to restrict the sample. 'Children' was used as part of each phrase in an attempt to locate websites that were child-specific.

Each search phrase was entered into each search engine (a total of nine searches) and the first 10 website links were selected from each search (a total of 90 website links). If the same website link appeared for a second time, it was discarded and the next non-repeated website link was included as part of the study. The search was limited to English language sites.

If a website link was not weight-related, it was still counted as a website link, but was not further analysed. A website link that was weight-related, but targeted to a group other then children, was still evaluated as part of the study. If a website link was weight-related but contained no information in the site, it was not further analysed. All selected website links were evaluated by one author (AO).

The base information collected on each website included target group, web link form, site type, country of origin, institutional origin and author(s) information. The website was also identified by the type of weight management programme that it provided, and the accessibility of the information or service.

The content information areas were the definition and aetiology of obesity, medical and psychosocial complications of obesity, dietary management, increasing physical activity, behaviour modification, parental and family involvement and the need for long-term management of obesity. A total of 28 clinical criteria were specified for the content areas.

Websites were rated according to the level of coverage for each of the 28 clinical criteria. Each criterion was given one of three rankings: 'not addressed', 'minimally addressed' or 'detail supplied'. Each criterion that was 'minimally addressed' or 'detail supplied' was analysed for its health information accuracy: Mostly inaccurate, Generally accurate and Fully accurate. (13) The Fry readability graph (FRG) method was used to assess the readability level of each website. (14,15) The average sentence length and the average number of syllables were calculated in the first page with a paragraph of more than 100 words and plotted on the FRG to provide a reading grade level--primary, high school, undergraduate or postgraduate.

The data were analysed using SPSS Version 10 (SPSS, Chicago, IL, USA), Minitab (Minitab Inc., State College, PA, USA) and Microsoft Excel 2000. To determine if a statistical significance occurred in the results between different search engines, a General Linear Model was performed using SPSS version 10. This was denoted by a P-value of less than 0.05. ANOVA was used to compare clinical criteria between website type.

RESULTS

A General Linear Model indicated that there was no significant difference between the search engines in regard to the individual variables (data not shown). The total number of sites generated from all the searches is shown in Table 1. Of the 90 websites generated for analysis, 66.7% (n = 60) were evaluated whereas 33.3% (n = 30) were unrelated to the issue of obesity. The results of the tested variables generated from the searches were shown to have no significant differences between each search engine, indicating some uniformity in the system.

Of the 60 websites so generated, 13 websites were specific for child obesity and the other 47 websites were on obesity in general. Of the websites 90% (n = 54) were North American, three were from the UK, two were Asian and one was European. The websites that were specific to childhood obesity were from the USA (n = 12) and Asia (n = 1). Of the 13 websites specific to childhood obesity, 10 were in the form of a website, whereas three provided information in the form of a journal article. No Australian websites were identified using the initial search criteria.

When the term '... in Australia' was added to the search somewhat different search results were yielded, including six Australian sites for hospital-based services, guidelines and journal articles, two of which were unable to be accessed. The four sites accessible at the time of the study were http://www.mja.com.au/public/nutrition/contents.html, http://www.phaa.net.au/policy/obesity.htm, http://www.cnw.au/parents/factsheets/weight_management_tips.htm and http://www.healthinsite.gov.au

The overall content of the sites was characterised as information, advice or sales and promotion. From the 60 websites analysed, 14 websites provided information only, 10 had both information and advice, 19 websites were involved in direct sales of information and products, and 17 in the promotion of organisations, services and products. Sale items included vitamins and minerals, meal substitutes and over the counter weight management medication. No website relating specifically to childhood obesity was involved in sales. Three childhood websites provided information; five supplied advice and five promoted a service.

Three of the child-specific websites had medical authors. Ten child websites had multidisciplinary health professional authorship. Five community-based childhood obesity websites were identified, which provided information for the purpose of awareness and education. There were five commercial websites that were specific to childhood obesity management and where programmes came at a cost. These sites included http://www.committed-to-kids.com (Committed to Kids Pediatric Weight Management Program) and http://www.shapedown.com (Shapedown Program). The other three childhood-specific obesity sites were hospital- (1) and university-based (2) programmes. Of the websites specific to childhood obesity six had no access costs, three required a fee to access their information, and four websites required paid attendance at a service to access treatment. Fees varied between $US 3.00 and $US 85.00 when quoted.

The 47 identified websites that were not specific to childhood obesity were categorised as commercial (30), hospital-based (7), university-based (3) and community (7). Because of the search conditions these websites all mentioned child obesity. In the commercial category there was no information regarding suitability or otherwise for children.

On the Fry readability system the majority of the 60 websites required a reading ability of an undergraduate (n = 26) or high school level (n = 18). Of the eb sites 15 had a reading grade equivalent to a postgraduate level and only one had a primary school level reading level.

The majority of the 60 assessed websites did not cover all management aspects with a mean value for 'not addressed' of 83.4% (range 95-65%). Of the websites that had 'minimally addressed' and 'detail supplied' for the selected management criteria, the majority had content that was completely accurate. In total, 15 websites defined childhood obesity as a body mass index (BMI) based on age-related centile charts, with three of these having generally accurate and 11 fully accurate content. Only three of the 13 websites specific to childhood obesity management addressed age-related BMI centile charts.

Genes and the environment were identified as aetiological factors of obesity in 14 and 15 websites, respectively, with 10 of these websites citing the aetiology of obesity as both genetic and environmental. The health risks of childhood obesity, when addressed, provided accurate content. Only one website addressed all three levels of morbidity criteria, three websites addressed two levels of health risks, one website discussed current major morbidities, and a total of 17 websites only acknowledged future medical problems as a relevant health risk. The websites that identified psychosocial morbidity in childhood obesity were fully accurate in content for both low self-esteem (n = 13) and social isolation (n = 8). Eight of the websites (13%) identified weight maintenance as an appropriate therapy goal for childhood obesity, with weight reduction necessary only in more extreme obesity cases.

Eleven sites discussed the food pyramid and five addressed dietary guidelines, with accurate content. Recommended daily dietary intakes were discussed in four websites. Four websites provided accurate information on food labelling, and the 10 websites that provided information on healthy snacks contained fully accurate and useful information. An explanation of the role of dietary fat in promoting obesity and a discussion on fat-containing foods was provided by 10 websites (17%). The recommended daily fat consumption varied widely between websites.

Five of the 60 websites (8%) discussed the different types of physical activity, sedentary behaviours and situations that prevent physical activity. Affirmation of the beneficial effects of physical activity and advice on how to increase physical activity appeared in 11 and 18 sites, respectively.

Of the 60 websites 13 addressed modification of eating behaviour and 12 addressed behaviour modification of physical activity. A total of 10 sites considered how to provide this information in an age-appropriate manner. Of the websites 10% contained all three criteria related to behaviour modification.

The importance of parental involvement in the management of childhood obesity was described in 11 websites and family participation in lifestyle changes advocated in 15. This information was graded as fully accurate in both the general and child-specific websites. Of the 60 websites 10 made mention of the benefits of or recommended long-term intervention for obesity management.

Statistically the child-specific sites were superior to the general sites in content coverage. Because of the low overall frequency of 'minimally addressed' and 'detail supplied' an overall P-value was calculated (P = 0.000), rather than a P-value for each criterion. The P-value for an individual criterion was however, statistically significant for the acknowledgement of low self-esteem in childhood obesity, the social isolation of obesity, and increasing activity and reducing sedentary behaviours in the management of obesity.

DISCUSSION

Consumers use the Internet as a tool to access health-related information. The present study investigated the accessibility, information quality and readability of web-based information for the management of overweight in children, in order to get some idea of how useful such sites might be to the interested parent. The search was made as if the parent of the obese child was performing the task (accessibility and readability), and the quality of the information was assessed from the health professional's viewpoint. The main findings of the present study are that little child-specific information is easily available, that this information when available was generally accurate and that a fairly sophisticated reading ability level was required to understand the information provided.

Searching the Internet is a somewhat serendipitous process for most consumers. An assumption was made that parents would limit the number of sites that they visited under a search condition. The choice of search procedure was clearly an arbitrary one, as was the choice of search phrase. However, the search procedure was logical, based on how search engines work. Specific guides that contain information for locating web-based health information are not readily available. The search used to generate search links was only moderately efficient, with one-third discarded as not relevant to the present study. There were only a few sites specific to children and obesity management. This finding suggests that either websites are difficult to locate through search engines and phrases, or that there are few websites specific to that subject. The latter seems probable given that there is a paucity of quality published data on the management of childhood obesity. (6) The specific child obesity sites provided both information and practical advice, and overall were of higher quality in terms of information provided than the general obesity sites. The Committed to Kids and Shapedown programs have both published their data. (16,17) Assuming that the number of child-specific obesity websites is truly low, then the probability that parents will effectively access these sites is also low and an opportunity to provide for informed parental action is lost. Most childhood obesity websites were of North American origin. A significant body of research on childhood obesity has come from non-English-speaking countries (18-20) and childhood obesity sites are generally institutional. Therefore, it is likely that the present study missed websites of potential interest but linguistically unavailable. The utility of specifying a country in the search phrase was illustrated by using the term '... in Australia' at the end of the search phrases, which generated sites specific to Australian organisations. Although no additional useable sites were generated, contacts were provided that a determined parent could pursue.

The child-specific obesity websites generally did not address all or most of the selected topics and clinical criteria, but when topics were addressed the content was accurate. This finding is similar to that of Berland et al. (13) The present study reported that the coverage of several key issues related to the topic of obesity was inconsistent and patchy, but that content accuracy itself was high. Important content that was often missing from child obesity sites included how to interpret the BMI in children, the occurrence of obesity-related medical morbidity in childhood, information on major sources of dietary fat and advice on increasing physical activity. These omissions are important, as knowledge around morbidity and physical activity may be an important influence on parental decisions about the management of their child's obesity. (21)

The FRG method was selected for its simplicity and for its previous application in published website analysis. (13) The reading grade level of websites required a high reading ability to comprehend the relevant content. It would seem reasonable for such websites to have a reading level of a primary to an early high school grade to ensure some equity of access.

Reservations about the reliability of health information in websites must exist. (22,23) Information should be complementary to, but not replace clinical management. In child obesity management it is recognised that clinical presentation occurs late, (24) with parents potentially seeking other avenues of advice. It is, therefore, possible that they will seek Internet information. Some health-based information sites on the Internet are now using e-health codes and guidelines. Relevant sites include Internet Healthcare Coalition http://www.ihealthcoalition.org/ethics/ethics.html, Health On the Net Foundation http://www.hon.ch/HONcode/Conduct.html, Netscoring http://www.chu-rouen.fr/netscoring/netscoringeng.html and Medcertain http://www.medcertain.org There are also publications discussing the issues to consider in looking for quality websites containing health information for patients. (25,26) Consumers are advised to consider both the source of the information, and any possible bias in the information provided. Suggestions to aid in recognising if the group behind the site are experts in the field are provided. The JAMA information sheet emphasises that the source of information should be easily found, that the authors should be clearly identified, references listed and an expert's review of the content, stating that it was accurate and complete at that time, be available. (25)

CONCLUSION

The Internet has the power to inform, and can be a valuable source of information. The management of childhood obesity is a health priority. Institutions and/or groups who are perceived by the lay public as being authoritative on the management of childhood obesity might consider their role as producers of quality Internet information. Such websites need to be easy-to-read and to provide practical help or direction for parents, on where to go, what to do and when to obtain professional assistance. At present there is a lack of quality, accessible information on the Internet for parents about childhood obesity and its management. Although the optimal management for childhood obesity should involve health professional input, there are limited resources for childhood obesity management in the health sector. In addition, not all families choose to seek professional help, and access to good quality information on the Internet is highly desirable.

ACKNOWLEDGEMENTS

The present study was carried out by AO as part of the requirements of a Master of Nutrition and Dietetics, Human Nutrition Unit, University of Sydney. The authors wish to thank the School of Mathematical Science, University of Technology, Sydney for assistance with statistical analysis.

REFERENCES

1 Lobstein T, Baur L, Uauy R for the IASO International Obesity Taskforce. Obesity in children and young people: a crisis in public health. Obes Rev 2004; 5 (Suppl. 1): 4-104.

2 Booth ML, Wake M, Armstrong T, Chey T, Hesketh K, Mathur S. The epidemiology of overweight and obesity among Australian children and adolescents, 1995-97. Aust N Z J Public Health 2001; 25: 162-9.

3 Lazarus R, Wake M, Hesketh K, Waters E. Change in body mass index in Australian primary school children, 1985-1997. Int J Obes Relat Metab Disord 2000; 24: 679-84.

4 Guo SS, Huang C, Maynard LM et al. Body mass index during childhood, adolescence and young adulthood in relation to adult overweight and adiposity: the Fels longitudinal study. Int J Obes Relat Metab Dis 2000; 24: 1628-35.

5 French SA, Story M, Perry C. Self-esteem and obesity in children and adolescents: a literature review. Obes Res 1995; 3: 479-90.

6 Summerbell CD, Ashton V, Campbell KJ, Edmunds L, Kelly S, Waters E. Interventions for treating obesity in children. The Cochrane Library Oxford. Protocol first published Issue 4, 1999; Review first published Issue 3, 2003.

7 Jadad AR, Gagliardi A. Rating health information on the Internet: navigating to knowledge or to babel? JAMA 1998; 279: 611-14.

8 Fox S, Raine L. The Online Health Care Revolution: How the Web Helps Americans Take Better Care of Themselves. Washington, DC: Pew Charitable Trusts, 2000.

9 Impicciatore P, Pandolfini C, Casella N, Bonati M. Reliability of health information for the public on the World Wide Web: systematic survey of advice on managing fever in children at home. BMJ 1997; 314: 1875-9.

10 Silberg WM, Lundberg GD, Musacchio RA. Assessing, controlling and assuring the quality of medical information on the Internet: Caveant lector et viewor--let the reader and viewer beware. JAMA 1997; 277: 1244-5.

11 Eysenbach G, Powell J, Kuss O, Sa E-R. Empirical studies assessing the quality of health information for consumers on the World Wide Web; a systematic review. JAMA 2002; 287: 2691-700.

12 Rippen H, Risk A. e-Health code of ethics. J Med Internet Res 2000; 2 (Suppl. 2): e9. (Cited 1 March 2006.) Available from URL: http://www.jmir.org/2000/2/e9/

13 Berland GK, Elliott MN, Morales LS et al. Health information on the Internet: accessibility, quality and readability in English and Spanish. JAMA 2001; 285: 2612-21.

14 Fry E. A readability formula that saves time. J Reading 1968; 11: 513-16, 575-8.

15 Fry E. Fry's readability graph: clarifications, validity and extension to level 17. J Reading 1977; 21: 242-52.

16 Sothern MS, von Almen TK, Schumacher HD, Suskind RM, Blecker U.A multidisciplinary approach to the treatment of childhood obesity. Del Med J 1999; 71: 255-61.

17 Mellin LM, Slinkard L, Irwin CE. Adolescent obesity intervention: validation of the SHAPEDOWN program. J Am Diet Assoc 1987; 8: 333-8.

18 Nuutinen O, Knip M. Long term weight control in obese children: persistence of treatment outcomes and metabolic change. Int J Obes 1992; 16: 279-87.

19 Maffeis C, Schutz Y, Pinelli L. Postprandial thermogenesis in obese children before and after weight reduction. Eur J Clin Nutr 1992; 46: 577-83.

20 Braet C, Van Winckel M, Van Leeuwen K. Follow up results of different treatment programs for obese children. Acta Paediatr 1997; 86: 397-402.

21 Epstein LH, Paluch RA, Gordy CC, Darn J. Decreasing sedentary behaviors in treating pediatric obesity. Arch Pediatr Adolesc Med 2000; 154: 220-26.

22 Kahn CE. Design and implementation of an Internet-based health information resource. Comput Methods Programs Biomed 2000; 63: 85-97.

23 Darmoni Sj, Haugh MC, Lukacs B, Boissel JP. Quality of health information about depression on Internet: level of evidence should be gold standard. BMJ 2001; 322: 1367.

24 Etelson D, Brand DA, Patrick PA, Shirali A. Childhood obesity: do parents recognize this health risk? Obes Res 2003; 11: 1362-8.

25 JAMA Patient Page. Health information on the Internet. JAMA 2001; 285: 2672.

26 Kiley R. Finding health information on the Internet: health consumers. Hosp Med 2000; 61: 799-801.

Amal OWAIMRIN (1) and Katharine STEINBECK (1,2)

(1) Metabolism and Obesity Services, Department of Endocrinology, Royal Prince Alfred Hospital, and (2) Faculty of Medicine, University of Sydney, Sydney, New South Wales, Australia

Correspondence: K. Steinbeck, Metabolism and Obesity Services, Royal Prince Alfred Hospital, Missenden Road, Camperdown, NSW 2050, Australia. Email: [e-mail]

A. Owaimrin, APD, Project Coordinator, Clinical Dietitian 

K. Steinbeck, PhD, MB BS, FRACP, Senior Staff Specialist

After this study was carried out the NHMRC Clinical Practice Guidelines for the Management of Overweight and Obesity in Children and Adolescents were published and made available on the Internet at URL: http://www.obesityguidelines.gov.au. This website now appears under the specified search conditions.

Table 1 Number of websites generated from the searches

                           Search engine
Search phrase (a)  AltaVista   Google   Excite

1                  39 291 293  627 000  2 907 620
2                  64 568 830  485 000  4 725 665
3                  42 183 970  579 000  2 817 890

(a) Search phrases: 1 = weight loss and children; 2 = weight management
and children; 3 = weight treatment and children.